Talk:Calculating Magic Damage
Credit goes to Aurik for translation of the Japanese website. I figured since we have pages for calculating WS and physical damage, we may as well have one for magic damage too. -Nivlakian 09:51, 27 August 2006 (EDT) Also of note... The Japanese convention for MAB and MDB in the source is different from the convention used on wikipedia. For example, the Japanese site writes MDB200. Most would consider MDB to start at 1 and write that as MDB+100 or say it has 100 MDB. -Nivlakian 10:08, 27 August 2006 (EDT) After some testing, it appears that AM2 also has an M of 2.0 and the same constants (V) of 710. -Nivlakian 13:44, 27 August 2006 (EDT) Re: Sorc Gloves / Static earring, the JP website seemed to indicate they didn't multiply by 1.05 but rather added 0.05 to the modifier. Is this an intentional change from my translation? --Aurikasura 16:32, 27 August 2006 (EDT) Actually, when I was rewording that part, I overlooked your wording and ended up writing it according to my personal observations. For example, pre-nukenerf, I nuked the Limbus version of Fafnir intending to MB w/ gloves/earring, but the SC did not go off. My thunder IV landed for 1060 damage. In that same setup, I nuked again to MB and it resulted in 1514 damage which is 42.8% more damage. The closest I can figure to get that number is 1.3 (MB) x 1.1 (equip). My friend tested his gloves before, but I don't have his numbers, but it was a 5% increase. Adding 5% to 30% to become 35% results in a ~3.8% increase over previous damage which was not what we experienced. I can revert it to your version if you'd prefer or perhaps should be marked for further confirmation? -Nivlakian 20:18, 27 August 2006 (EDT) Mark it for further investigation, I suppose. I have neither so I can't really confirm it :P --Aurikasura 21:11, 27 August 2006 (EDT) I've posted MB tests in my usertalk. Interestingly, additional SCs does not add 20% damage but instead 5%. Also, it appears the MB damage is calculated prior to MAB. ie. (D*resist)*MB*MBGloves*MBEarring*MAB with flooring after each part is multiplied. Credit goes to Suiram of Pandemonium for analyzing the results. I'm going to alter the page accordingly. - Nivlakian 15:09, 29 August 2006 (EDT) Awesome data. Thanks! --Aurikasura 15:27, 29 August 2006 (EDT) ---- Great job there to everyone involved in compiling that. I've wondered about additional SCs' effect on MB bonus for a while. --Ichthyos 15:10, 29 August 2006 (EDT) ---- Suiram of Pandemonium has done some testing regarding the effects of triple ice in Beaucedine Glacier on Tundra Tigers and finds that it is not capped at 1.3, but will go to 1.35. He used the 100INT and 1.5MAB for the testing as was done with some of our other tests. What was interesting is that he has found that Sorcerer's Tonban has no effect during triple conditions suggesting that the term is capped at 1.35. I will modify the information accordingly. -Nivlakian 14:38, 14 September 2006 (EDT) ---- A bit more info for magic damage reduction While I was translating Blue Magic Damage page, I asked Apple Pie in DiV a few question regarding certain variables, and he pointed me out a Good JP BLM website. In that website, There's additional information about magic damage cut from different spells/monster JA/family I've translated that section in Calculating Weapon Skill Damage page. I hope we can synchronize this page and other damage calculation pages. Another thing is the detail of the calculation. According to that site, we should do truncation for each step of multiplication. That is: ...int(int(int(var1 x var 2)*var3)*var4) Apple also pointed me out this website for cap on each BLM spells. I notice there was inconsistency between numbers supplied by Sloppysue and the ones provided on that site. Which one should we follow? don't ask me. I would translate them, but I kinda don't have time right now--VZX 05:52, 22 October 2006 (EDT) ----- The "Good JP BLM website" you referenced is already listed among the sources at the bottom. From further testing, it seems some of the information on there was incorrect and most of those corrections are mentioned above in the discussion. Also, flooring at each step is mentioned at the top of the page: "Actual Damage is calculated by multiplying the following factors in order and flooring after each step. " It is written that way since a long string of int( or floor( becomes difficult to read. -Nivlakian 01:48, 24 October 2006 (EDT) Shell This doesn't touch on how the Shell line of spells reduce magic damage. It seems that they are applied separately from MDB, but I haven't seen a formula anywhere. i.e. is Shell IV -22% or 200/256 or something else?--Valyana 02:42, 25 October 2006 (EDT) ---- According to this reference, Shells, Shining Ruby, Bubble curtain are categorized as "Magic Damage Cut". This is not the same with Target Magic Damage Reduction section though (as far as I comprehend the JP site). Should we change the definition? I have translated the lists of Magic Damage Cut buff in magical WS damage section though--VZX 06:08, 25 October 2006 (EDT) ---- In Aurik's original translation of that site, the section with information on Shell is what became "Target Magic Damage Reduction". However, he did not include that data in the translation. There are also some discrepancies between the values listed in that source and people's personal observations, and given the number of inaccuracies already discovered there, it is probably best to omit them until they can be confirmed. Suiram 19:23, 25 October 2006 (EDT) ---- I think this line Note: most HNMs don't have this effect, but rather a huge MDB value (ex. Tiamat = MDB+100, Jormungand = MDB+124) should go into Magic Attack Bonus / Magic Defense Bonus section, not Target Magic Damage Reduction ----- I added a section on the reduction in damage when casting on multiple targets with -ga spell, and inserted it into the appropriate part of the calculation. I'll amend my user talk page with the details of the tests used to arrive at those values and its location in the overall formula in the coming days. I also adjusted the resist section to remove 0 as a possible value in that part of the equation. As for the line regarding some HNMs not possessing Target Magic Damage Reduction, I think it's fine where it is. -Suiram 02:29, 26 October 2006 (EDT) ----- I added "with respect to magic bursts" to the sentence "Further testing would be necessary to confirm the order of the other parts." The order of the other portions staff, day/weather, etc have already been tested. -Nivlakian 15:58, 26 October 2006 (EDT) MDB I just want to ask to confirm something: :*Magic Attack Bonus / Magic Defense Bonus :*MAB is divided by MDB. It says it's divided by instead of subtracted from then multiplied to the magic damage. Using a previous example to state my question, if a 75BLM cast Thunder IV on Tiamat, with MDB+100, would the overall damage mutiplication be .32 (MADIV - 100) or would it be .66? (1.32 / 2)--Dragonspight 05:16, 22 November 2006 (EST) ----- The overall damage multiplier--if this section is correct--would be 1.32/2.0 = 0.66. However, both the statement that MAB/MDB takes place without intermediate flooring and the statement that Tiamat has 2.0 MDB rather than 50% Target Magic Damage Reduction are taken directly from the original translation and are as-yet unconfirmed aspects of the Magic Damage formula, so take it with a grain of salt. However, even if those parts are inaccurate, it would still work out to roughly 0.66. Suiram 16:00, 22 November 2006 (EST) M and V I recently completed tests to confirm the M and V values for all regular single-target series, -ga series, and AM series I nukes. All of the M values were as previously documented, but the values for V previously listed, which came from allakhazam source, were incorrect for most I and II series nukes as well as often missing and sometimes incorrect for -ga spells. I didn't test AM II M or V because those were recently tested by Nivlakian and I trust his methodology. I'll post the data from my tests on my Usertalk page sometime in the next few days. I also made some minor cosmetic changes to that section, such as changing the table format so there were borders between cells. -Suiram 19:25, 22 November 2006 (EST) Target Magic Damage Reduction I renamed this section to say "Adjustment" since it can be either a bonus or reduction to the damage a particular monster takes. Previously, the heading makes it seem that it can only be a number less than 100%. -Nivlakian 19:43, 22 November 2006 (EST) Ninjutsu, divine magic, and Others Would some lovely soul be willing to go about finding the damages for Ninjutsu nukes and other damage spells so that they can be applied to this formula~? --Dragonspight 09:40, 2 December 2006 (EST) Reverted Changes If you're going to change something in this article, please at least present some supporting evidence for the change. I have reverted the article to the previous version. -Nivlakian 13:02, 16 December 2006 (EST) Incorrect Magic burst and Weather bonus Double weather is 1.2, not 1.25 Magic burst is always 1.33 unless you're wearing Static Earring and/or Sorc. Gloves. These ADD, not MULTIPLY, 0.05 to the multiplier. The number of WS in the renkei does not change anything. I made these corrections and they were reverted... why? Would you like a test with screenshots or what? Also: Actually, when I was rewording that part, I overlooked your wording and ended up writing it according to my personal observations. For example, pre-nukenerf, I nuked the Limbus version of Fafnir intending to MB w/ gloves/earring, but the SC did not go off. My thunder IV landed for 1060 damage. In that same setup, I nuked again to MB and it resulted in 1514 damage which is 42.8% more damage. The closest I can figure to get that number is 1.3 (MB) x 1.1 (equip). 1060 * (1.33+0.10) = 1515 Your damage didn't add up correctly because the multiplier is 1.33, not 1.3 -Kerberoz 15:52, 19 December 2006 (EST) If you could provide screenshots or other logs along the lines of what can be found on Niv's talk page (http://wiki.ffxiclopedia.org/User_talk:Nivlakian) it would be easier to confirm your 3 assertions: 1) weather x2 is 1.2 weather x2 + day is +0.35, proven from Suiram's test day is +0.1 therefore we conclude that weather x2 is +0.2 2) base MB modifier is 1.33, not 1.3 This is drawn from Niv's MB test Normal thunder2 damage: 343 2-part Light MB thunder2 damage: 445 * 1.3 = 445 * 1.33 = 456 Clearly 1.30 gives a better approximation than 1.33. 3) Sorc gloves / static earring add 0.05 each instead of multiplying by 1.05 individually / 1.10 paired This is drawn from Niv's MB test Static Earring or Sorc Gloves only Normal thunder2 damage: 343 2-part Light MB thunder2 damage: 466 Guessing at the formula... * 1.35 = 463 * 1.38 = 473 [* 1.3 * 1.05] = 467 [* 1.3 * 268/256] = 465 [* 1.05 * 1.3] = 468.0 [* 268/256 * 1.30] = 466 Static Earring + Sorcerer's Gloves Normal thunder2 damage: 343 2-part Light MB thunder2 damage: 489 * 1.40 = 480 * 1.43 = 490 [* 1.3 * 1.10] = 489 [[* 1.3 * 1.05] * 1.05] = 490 [* 1.3 * 280/256] = 486 [[* 1.3 * 268/256] * 268/256] = 486 [* 1.1 * 1.3] = 490 [* 280/256 * 1.3] = 487 [[* 268/256 * 268/256] * 1.3] = 487 etc... I can't find the exact formula used, but the closest we've found is [V*mb*mult], where mb is 1.3, 1.35 etc and multi is 1.05-1.10 depending on gear used. --Aurikasura 20:36, 19 December 2006 (EST) ---- I was the Hume BLM in those MB tests, and they ended up revealing some very important information in a few ways for the damage formula. The JP wiki that serves as the foundation for this page that Aurik originally translated and which is sourced at the bottom had a number of inaccuracies with respect to the Magic Burst section. For one, it listed each additional renkei as adding 0.2 to the MB bonus, which we confirmed to be only 0.05. There was also a question as to whether or not AF2 hands and static earring were added to the MB term or multiplied separately, which the tests confirmed for the latter. Finally--and I believe most importantly due to its subtlety--the data we obtained showed that the MB term could not be calculated after the MAB term and still arrive at the final damage values over all of the tests due to flooring. Anyhow, here is the relevant portion of the formula for those tests (dINT=34, M=1.5, V=178, MAB=1.5 for all): floor(floor(floor(floor(229 * MB) * MB1) * MB2) * 1.5) MB is the MB bonus, MB1 and MB2 are the respective bonuses from AF2 hands and static earring. If you plug in the values described by the current formula you see that it works out to exactly the damage we observed. If you happen to find some other values that would also arrive at the same damage in every case, then feel free to develop a test that would determine which of the models is correct and edit this page in accordance with your findings. Please don't, however, edit in information that you erroneously believe to be the case simply on the basis of your unfounded beliefs. -- Suiram 05:43, 20 December 2006 (EST) ---- To help clarify how to calculate the magic burst damage section, I added an example. I have also added examples of Target Magic Damage Adjustment. The Ebony Pudding and Eo'ghrah information come from my own testing while the others are cited in the Japanese source at the bottom of the page. One thing that I'm curious about is whether Target Magic Damage Adjustment and Shell are cumulative or multiplied seperately. While solo'ing a light Aw'ghrah once a long time ago, my Thunder IV did 964 damage. The Aw'ghrah then cast Shell IV on itself and Thunder IV then did 723 which is exactly 75% of previous damage. Assuming Aw'ghrah have a magic damage adjustment similar to Eo'ghrah (or even the same), it appears they may be multiplied seperately in order to arrive at such a clearcut 25% loss of damage. Although I have screenshots of this, unfortunately, I don't recall my exact setup at the time making it impossible to do a proper/thorough calculation. I may try to test it again sometime. -Nivlakian 18:17, 3 January 2007 (EST) Actually, thinking about it more and using the value for Shell IV in the JP source, the magic adjustment is probably cumulative. Assuming Aw'ghrah share the 87.5% reduction of Eo'ghrah, the Thunder IV damage prior to reduction could have been 1102 as 1102 * (1- 32/256) = 964.25. Shell IV is supposed to reduce damage by 56/256... (1102 * (1 - 32/256 - 56/256)) = 723.1875. This properly explains the damage. -Nivlakian 18:32, 3 January 2007 (EST) ---- I removed the line about Tiamat and Jormungand altogether, as preliminary tests indicate that it's neither strictly MDB as previously entered nor an exact 50% damage TMDA reduction. I also re-posted the link to the the BG thread that serves as the foundation for this page, but at the new bluegartrls.com domain. Ebony Puddings take around 10% more magic damage, not 25% as stated on the page. I'm not sure of the exact value but it is certainly not 25%. If you consult Nivlakian's usertalk page, you can verify for yourself the tests he used to determine the INT and TMDA for Ebony Pudding. -Suiram 11:56, 22 February 2007 (EST) Tests Regarding Double Weather Conditions As this continues to be a point of confusion, I'm reposting the data here. The day/weather bonus for double-weather was originally translated as +0.25. Since this went against the general understanding of +0.2, I set out to verify it. An image of the chatlog for the relevant portion of my test can be found here: http://img170.imageshack.us/img170/8832/img0051ik0.jpg In this chatlog entry, you can see me casting on three different tigers. For every cast I had 100 INT and 1.5 total MAB, with triple-ice weather present. The first tiger I did not wear sorcerer's tonban at all, the second tiger I wore sorcerer's tonban, and the third I did not wear it for the first cast, but wore it for the second cast. floor(floor(dINT*2+541)*1.5) = 1030 solves for dINT = 73 floor(floor(floor(73*1.5 + 320)*1.35)*1.5) = 868 So the bonus for day and double weather with no sorcerer's tonban is 1.35. Since Day bonus is known to be +0.1, both from previous tests and an earlier part of above test, this suggests that the bonus for double weather is indeed +0.25. The damage to the second tiger suggests that the term is capped at 1.35 (whereas it should otherwise be 1.4 with day, double weather, and sorcerer's tonban), and the the third tiger is simply verification of the existence of a cap. -Suiram 03:53, 1 March 2007 (EST) I adjusted the line about the cap on this term to (hopefully) improve clarity and prevent future misunderstandings. -Suiram 12:23, 1 March 2007 (EST) Decimal Truncation I added a line regarding decimal truncation to both the MAB/MDB section and the TMDA section. Both of these claims will require further testing, but so far it is the only explanation I have to otherwise conflicting data. For MAB/MDB, I tested on an Aura Pot in the Shrine of Ru'Avitau, with no staff or day/weather bonus, and keeping my MAB constant at 1.5. I have independently determined Magic Pots to have a TMDA reduction of 50% using the Blue Magic spell 1000 Needles, which ignores MDB. With 100 INT Blizzard 4 and Thunder 4 did 368 and 391 respectively, while with 105 INT they did 375 and 398 respectively. Calculating MAB/MDB to an arbitrarily long decimal expansion with reasonable assumptions for the target's MDB resulted in each case in values that differed from the predicted values in at least one cast of the four, and similarly for a three-place truncation. The only assumption that made all of the values fit with their predictions was for a MDB value of 1.14, truncated to 2 decimal places, so that the final value of the term is 1.33. For TMDA, this is due to reports of the special weaponskill, Discoid, used by the Zhayolm Remnants boss, Battleclad Chariot. When used against a single player target with no TMDA, it does 4400 damage, however when used against a single player target with Shell 4, it does 3436 damage. Since floor(4400 * (200/256)) = 3437, floor(4400 * 0.78) = 3432, and floor(4400 * 0.781) = 3436, it seems reasonable at this time to conclude that this term also undergoes decimal truncation before being calculated in the overall damage formula. -Suiram 08:15, 29 March 2007 (EDT)